OBIS Marine Biodata Workshop

Overview

Notebook developed with R version 4.2.2

Use Case

• Interoperable IFCB data product for the CA HAB Bulletin
• IFCB Dashboard to OBIS use case with automated classification
• Event core with Occurrence extension

Authors

Ian Brunjes (SCCOOS), Stace Beaulieu (WHOI) Prepared for OBIS IOOS Marine Biological Data Mobilization Workshop April 2023

This is a prototype for testing purposes only. A protocol is being developed to determine if and when appropriate to submit products from automated classification to OBIS.

Sponsored by NOAA PCMHAB20 project “Harmful Algal Bloom Community Technology Accelerator”

Workflow

Main steps in the workflow can be related to the IFCB workflow diagram in the ifcb-analysis wiki on GitHub

Step: Classification - Interpretation for the autoclass scores / transform automated classification into presence/absence

Step: Matching class labels to scientific names and IDs - Match the class labels to scientific names in the World Register of Marine Species (WoRMS) taxonomic database.

Step: Summarization - Calculate concentration as number of ROIs classified to a taxon divided by volume analyzed

Next step (not shown / would extend the diagram): Transforming to Darwin Core - Map resulting data table into Darwin Core table(s)

Target data product to standardize to Darwin Core:

Concentration of 2 genera of HAB taxa from an IFCB sample(s) (e.g., here is a sample with autoclass available in HABDAC at Del Mar Mooring

Preconditions: - IFCB Dashboard sample (bin) has autoclass csv file with scores per class label from automated classifier - IFCB Dashboard sample has been populated with a dataset name, volume_analyzed, datetime, latitude, longitude, and depth - For autoclass labels: A lookup table has been prepared with thresholds per class label

This workflow is being developed to meet the EU Horizon 2020 “Best practices and recommendations for plankton imagery data management” http://dx.doi.org/10.25607/OBP-1742

Classification

In this step, we interpret the autoclass scores from the autoclass.csv file on the IFCB Dashboard. We will filter to the targeted class labels, apply a threshold per class label, and determine the “winning” class label per ROI, thus transforming the automated classification into a presence/absence table.

target_labels = read_csv(here("data", "target_classification_labels.csv"))

kable(target_labels) %>% kable_material(c("striped", "hover")) %>% scroll_box(width = "100%")

label	intended_worms_taxon	autoclass_threshold
Alexandrium catenella	Alexandrium	0.2
Pseudo-nitzschia	Pseudo-nitzschia	0.7
pennate Pseudo-nitzschia	Pseudo-nitzschia	0.7

The order of operations is important in this filtering and thresholding process. If the filtering is applied prior to thresholding, the concentration is possibly (likely) to be overestimated by excluding other classes that may have higher scores.

Thresholds used in this prototype are for testing purposes only.

Our initial prototype will retain as ‘absence’ (zero count) when no ROIs exceed per-class threshold. However, we acknowledge that data providers may want to use a different per-class threshold to report absence, and might only want to report presence.

Once the output from this Classification step is matched to scientific names and IDs (next step) the intermediate table loosely corresponds to Level 1b SeaBASS file (classification per ROI).

sample_bin = "D20210926T181303_IFCB158"
bin_details = get_bin_details(sample_bin)
# Thresholds used in this prototype are for testing purposes only.

bin_occurrences = get_bin_occurrences(sample_bin, target_labels)
kable(bin_occurrences) %>% kable_material(c("striped", "hover")) %>% scroll_box(width = "100%")

pid	class	score
D20210926T181303_IFCB158_01203	Pseudo-nitzschia	0.9565
D20210926T181303_IFCB158_01634	Pseudo-nitzschia	0.7026
D20210926T181303_IFCB158_01650	Pseudo-nitzschia	0.9560

Matching class labels to scientific names and IDs

In this step, we use (a portion of) each autoclass label to query the API of the World Register of Marine Species (WoRMS) taxonomic database to return an accepted scientific name, its paired Aphia ID, and its taxon rank and kingdom.

wm_records = get_worms_taxonomy(target_labels$intended_worms_taxon)
kable(wm_records) %>% kable_material(c("striped", "hover")) %>% scroll_box(width = "100%")

AphiaID	scientificname	lsid	rank	kingdom	intended_worms_taxon
109470	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	Alexandrium
149151	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	Pseudo-nitzschia

Summarization

In this step, we use the table from the Classification step and results from the Matching to WoRMS step to calculate concentration as number of ROIs classified to a taxon divided by volume analyzed per sample.

Output from the Summarization step loosely corresponds to Level 2 SeaBASS file.

occurrences_summary = summarize_bin_occurrences(bin_details, bin_occurrences, target_labels)
kable(occurrences_summary) %>% kable_material(c("striped", "hover")) %>% scroll_box(width = "100%")

intended_worms_taxon	occurrences	taxon_classes	associated_rois	occurrences_per_ml	sampleTime	lat	lng	bin_id
Alexandrium	0	Alexandrium catenella		0.0000000	2021-09-26T18:13:03+00:00	32.92917	-117.3165	D20210926T181303_IFCB158
Pseudo-nitzschia	3	Pseudo-nitzschia | pennate Pseudo-nitzschia	_01203 | _01634 | _01650	0.7109005	2021-09-26T18:13:03+00:00	32.92917	-117.3165	D20210926T181303_IFCB158

Transforming to Darwin Core

In this step, we transform the table from the Summarization step into two tables (an event table and an occurrence table) and add columns to meet OBIS and GBIF requirements for the Darwin Core Archive package.

We also add columns to meet the EU Horizon 2020 “Best practices and recommendations for plankton imagery data management” http://dx.doi.org/10.25607/OBP-1742

To assign the unique occurrenceID for the concentrations per taxon per sample we used eventID_taxonID, a pattern similar to the EU best practice, such that an individual included in the summed count would be represented by eventID_taxonID_roiID. Ultimately, we would like to test the DwC ResourceRelationship extension and/or DwC term associatedOrganisms to relate individuals to abundances reported to OBIS.

# Build the Darwin Core Event table
event_tbl = build_event_table(bin_details)
kable(event_tbl) %>% kable_material(c("striped", "hover")) %>% scroll_box(width = "100%")

datasetName	eventID	eventDate	decimalLongitude	decimalLatitude	countryCode	geodeticDatum	minimumDepthInMeters	maximumDepthInMeters	sampleSizeValue	sampleSizeUnit
del-mar-mooring	D20210926T181303_IFCB158	2021-09-26T18:13:03+00:00	-117.3165	32.92917	US	WGS84	0	0	4.22	milliliter

# Join occurrence summary with taxon records
wm_occurrences = left_join(occurrences_summary, wm_records, by = c("intended_worms_taxon" = "intended_worms_taxon"))

# Build the Darwin Core Occurrence table
occurrence_tbl = build_occurrence_table(wm_occurrences, bin_details)
kable(occurrence_tbl) %>% kable_material(c("striped", "hover")) %>% scroll_box(width = "100%")

eventID	occurrenceID	basisOfRecord	identifiedBy	identificationVerificationStatus	identificationReferences	identificationRemarks	associatedMedia	verbatimIdentification	scientificName	scientificNameID	taxonRank	kingdom	occurrenceStatus	organismQuantity	organismQuantityType	institutionCode
D20210926T181303_IFCB158	D20210926T181303_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210926T181303_IFCB158	D20210926T181303_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T181303_IFCB158&image=01203 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T181303_IFCB158&image=01634 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T181303_IFCB158&image=01650	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.7109005	counts per milliliter	AxiomROR

Extending workflow over a timespan

# Demonstrating how to build the DwC tables for all bins within a span of time
start_date = "2021-09-25"
end_date = "2021-09-27"

# Read in classification input parameters
target_labels = read_csv(here("data", "target_classification_labels.csv"))

# Query WORMS lookup table
wm_records = get_worms_taxonomy(target_labels$intended_worms_taxon)

# Get the ifcb bins within time span and construct
bin_ids = get_bins_in_range(start_date, end_date)

event_tables = list()
occurrence_tables = list()

for(bin in bin_ids) {
  if (bin_has_autoclass(bin)) {
    # Build occurrence summary
    bin_details = get_bin_details(bin)
    bin_occurrences = get_bin_occurrences(bin, target_labels)
    occurrences_summary = summarize_bin_occurrences(bin_details, bin_occurrences, target_labels)
    
    # Build dwc event table
    event_tbl = build_event_table(bin_details)
    
    # Build dwc occurrence table
    wm_occurrences = left_join(occurrences_summary, wm_records, by = c("intended_worms_taxon" = "intended_worms_taxon"))
    occurrence_tbl = build_occurrence_table(wm_occurrences, bin_details)
    
    event_tables[bin] = list(event_tbl)
    occurrence_tables[bin] = list(occurrence_tbl)
  }
}

# Bind each per bin result into single event/occurrence table
event_tbl = bind_rows(event_tables)
occurrence_tbl = bind_rows(occurrence_tables)

# Save to output to .csv
write_csv(event_tbl, here("output", "event.csv"))
write_csv(occurrence_tbl, here("output", "occurrence.csv"))

Aggregated event table

kable(event_tbl) %>% kable_material(c("striped", "hover")) %>% scroll_box(width = "100%", height = "400px")

datasetName	eventID	eventDate	decimalLongitude	decimalLatitude	countryCode	geodeticDatum	minimumDepthInMeters	maximumDepthInMeters	sampleSizeValue	sampleSizeUnit
del-mar-mooring	D20210926T211303_IFCB158	2021-09-26T21:13:03+00:00	-117.3165	32.92917	US	WGS84	0	0	4.336	milliliter
del-mar-mooring	D20210926T181303_IFCB158	2021-09-26T18:13:03+00:00	-117.3165	32.92917	US	WGS84	0	0	4.220	milliliter
del-mar-mooring	D20210926T151303_IFCB158	2021-09-26T15:13:03+00:00	-117.3165	32.92917	US	WGS84	0	0	4.482	milliliter
del-mar-mooring	D20210926T121304_IFCB158	2021-09-26T12:13:04+00:00	-117.3165	32.92917	US	WGS84	0	0	4.538	milliliter
del-mar-mooring	D20210926T091303_IFCB158	2021-09-26T09:13:03+00:00	-117.3165	32.92917	US	WGS84	0	0	4.578	milliliter
del-mar-mooring	D20210926T061303_IFCB158	2021-09-26T06:13:03+00:00	-117.3165	32.92917	US	WGS84	0	0	4.616	milliliter
del-mar-mooring	D20210926T031304_IFCB158	2021-09-26T03:13:04+00:00	-117.3165	32.92917	US	WGS84	0	0	4.562	milliliter
del-mar-mooring	D20210926T001304_IFCB158	2021-09-26T00:13:04+00:00	-117.3165	32.92917	US	WGS84	0	0	4.581	milliliter
del-mar-mooring	D20210925T211305_IFCB158	2021-09-25T21:13:05+00:00	-117.3165	32.92917	US	WGS84	0	0	4.328	milliliter
del-mar-mooring	D20210925T181304_IFCB158	2021-09-25T18:13:04+00:00	-117.3165	32.92917	US	WGS84	0	0	4.289	milliliter
del-mar-mooring	D20210925T151304_IFCB158	2021-09-25T15:13:04+00:00	-117.3165	32.92917	US	WGS84	0	0	4.166	milliliter
del-mar-mooring	D20210925T121306_IFCB158	2021-09-25T12:13:06+00:00	-117.3165	32.92917	US	WGS84	0	0	4.381	milliliter
del-mar-mooring	D20210925T091303_IFCB158	2021-09-25T09:13:03+00:00	-117.3165	32.92917	US	WGS84	0	0	4.554	milliliter
del-mar-mooring	D20210925T061303_IFCB158	2021-09-25T06:13:03+00:00	-117.3165	32.92917	US	WGS84	0	0	4.580	milliliter
del-mar-mooring	D20210925T031303_IFCB158	2021-09-25T03:13:03+00:00	-117.3165	32.92917	US	WGS84	0	0	4.558	milliliter
del-mar-mooring	D20210925T001302_IFCB158	2021-09-25T00:13:02+00:00	-117.3165	32.92917	US	WGS84	0	0	4.436	milliliter

Aggregated occurrence table

kable(occurrence_tbl) %>% kable_material(c("striped", "hover")) %>% scroll_box(width = "100%", height = "400px")

eventID	occurrenceID	basisOfRecord	identifiedBy	identificationVerificationStatus	identificationReferences	identificationRemarks	associatedMedia	verbatimIdentification	scientificName	scientificNameID	taxonRank	kingdom	occurrenceStatus	organismQuantity	organismQuantityType	institutionCode
D20210926T211303_IFCB158	D20210926T211303_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210926T211303_IFCB158	D20210926T211303_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T211303_IFCB158&image=00672 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T211303_IFCB158&image=00950	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.4612546	counts per milliliter	AxiomROR
D20210926T181303_IFCB158	D20210926T181303_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210926T181303_IFCB158	D20210926T181303_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T181303_IFCB158&image=01203 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T181303_IFCB158&image=01634 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T181303_IFCB158&image=01650	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.7109005	counts per milliliter	AxiomROR
D20210926T151303_IFCB158	D20210926T151303_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210926T151303_IFCB158	D20210926T151303_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T151303_IFCB158&image=00999	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.2231147	counts per milliliter	AxiomROR
D20210926T121304_IFCB158	D20210926T121304_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T121304_IFCB158&image=00076 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T121304_IFCB158&image=00928	Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	present	0.4407228	counts per milliliter	AxiomROR
D20210926T121304_IFCB158	D20210926T121304_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T121304_IFCB158&image=00739 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T121304_IFCB158&image=01033	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.4407228	counts per milliliter	AxiomROR
D20210926T091303_IFCB158	D20210926T091303_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210926T091303_IFCB158	D20210926T091303_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T091303_IFCB158&image=00133 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T091303_IFCB158&image=00188 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T091303_IFCB158&image=00519 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T091303_IFCB158&image=00556 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T091303_IFCB158&image=00575 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T091303_IFCB158&image=00673	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	1.3106160	counts per milliliter	AxiomROR
D20210926T061303_IFCB158	D20210926T061303_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210926T061303_IFCB158	D20210926T061303_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T061303_IFCB158&image=00481 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T061303_IFCB158&image=00882	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.4332756	counts per milliliter	AxiomROR
D20210926T031304_IFCB158	D20210926T031304_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210926T031304_IFCB158	D20210926T031304_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T031304_IFCB158&image=00188 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T031304_IFCB158&image=00982	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.4384042	counts per milliliter	AxiomROR
D20210926T001304_IFCB158	D20210926T001304_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210926T001304_IFCB158	D20210926T001304_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210926T001304_IFCB158&image=00694	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.2182929	counts per milliliter	AxiomROR
D20210925T211305_IFCB158	D20210925T211305_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210925T211305_IFCB158	D20210925T211305_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=00078 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=00110 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=00164 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=00287 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=00563 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=00717 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=00749 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=00845 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=00946 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=01037 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=01056 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=01220 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=01222 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=01256 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=01620 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T211305_IFCB158&image=01688	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	3.6968577	counts per milliliter	AxiomROR
D20210925T181304_IFCB158	D20210925T181304_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210925T181304_IFCB158	D20210925T181304_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T181304_IFCB158&image=00186 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T181304_IFCB158&image=00227 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T181304_IFCB158&image=00675 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T181304_IFCB158&image=00676 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T181304_IFCB158&image=01000 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T181304_IFCB158&image=01416 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T181304_IFCB158&image=01465 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T181304_IFCB158&image=01728 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T181304_IFCB158&image=01824	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	2.0983912	counts per milliliter	AxiomROR
D20210925T151304_IFCB158	D20210925T151304_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=01191	Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	present	0.2400384	counts per milliliter	AxiomROR
D20210925T151304_IFCB158	D20210925T151304_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=00125 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=00272 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=00364 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=00438 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=00606 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=00892 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=00905 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=00908 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=00952 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=01016 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=01201 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=01358 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T151304_IFCB158&image=01959	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	3.1204993	counts per milliliter	AxiomROR
D20210925T121306_IFCB158	D20210925T121306_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210925T121306_IFCB158	D20210925T121306_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T121306_IFCB158&image=00364 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T121306_IFCB158&image=00441 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T121306_IFCB158&image=00690 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T121306_IFCB158&image=00943 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T121306_IFCB158&image=01005 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T121306_IFCB158&image=01117 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T121306_IFCB158&image=01352	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	1.5978087	counts per milliliter	AxiomROR
D20210925T091303_IFCB158	D20210925T091303_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210925T091303_IFCB158	D20210925T091303_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T091303_IFCB158&image=00138 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T091303_IFCB158&image=00197 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T091303_IFCB158&image=00365	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.6587615	counts per milliliter	AxiomROR
D20210925T061303_IFCB158	D20210925T061303_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210925T061303_IFCB158	D20210925T061303_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T061303_IFCB158&image=01003	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.2183406	counts per milliliter	AxiomROR
D20210925T031303_IFCB158	D20210925T031303_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210925T031303_IFCB158	D20210925T031303_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T031303_IFCB158&image=00190 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T031303_IFCB158&image=00673 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T031303_IFCB158&image=01017 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T031303_IFCB158&image=01019	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.8775779	counts per milliliter	AxiomROR
D20210925T001302_IFCB158	D20210925T001302_IFCB158_109470	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.		Alexandrium catenella	Alexandrium	urn:lsid:marinespecies.org:taxname:109470	Genus	Chromista	absent	0.0000000	counts per milliliter	AxiomROR
D20210925T001302_IFCB158	D20210925T001302_IFCB158_149151	MachineObservation		PredictedByMachine	Trained machine learning model: 20220416_Delmar_NES_1.ptl (recommend publishing to a community or institutional repository for DOI) | Software to run the trained machine learning model: https://github.com/WHOIGit/ifcb_classifier (recommend referring to GitHub release or commit if not published for DOI) | Software to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/ifcb_autoclass_eval.R | Input parameters to interpret autoclass scores: https://github.com/sccoos/OBIS_workshop_2023_IFCB/blob/902b4fb504a552e0c25bbdee099495dbf2c1e143/data/target_classification_labels.csv	Arbitrary threshold used for both presence and absence without testing for false positives.	https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T001302_IFCB158&image=00368 | https://ifcb.caloos.org/image?dataset=del-mar-mooring&bin=D20210925T001302_IFCB158&image=01219	Pseudo-nitzschia | pennate Pseudo-nitzschia	Pseudo-nitzschia	urn:lsid:marinespecies.org:taxname:149151	Genus	Chromista	present	0.4508566	counts per milliliter	AxiomROR

Visualizing results

# Join event timestamp to occurrences
eo = left_join(occurrence_tbl, event_tbl, by = join_by(eventID)) %>%
  mutate(eventDate = as_datetime(eventDate)) %>%
  select(eventID, eventDate, scientificName, organismQuantity)


# Built plotly
p = eo %>% ggplot(aes(x=eventDate, y=organismQuantity, color = scientificName)) + geom_point() + geom_line() + labs(x = "", y = "organismQuantity (counts per milliliter)", title = paste0("Del Mar IFCB Occurrence Data (", start_date, " - ", end_date, ")"))

ggplotly(p) %>% layout(legend = list(title = "", x = 0.75, y = 0.9))